HR leaders have a choice about whether to get ahead of this view or manage the wreckage
Brian Chesky does not deal in euphemism. The Airbnb co-founder and chief executive, speaking on the Invest Like the Best podcast earlier this month, identified two categories of worker he believes will not survive the transition to an AI-enabled workplace: people who refuse to adapt, and people managers whose entire contribution is managing people.
"I do not think people managers will have any value in the future," Chesky said. "When I mean people managers, I mean people that only manage people. You cannot just be these managers where you are people's therapists, and you are just doing meetings, you are doing one-on-ones."
The remark was widely reported - because, for many people working in or alongside corporate management structures, it named something already being felt. The layer of management that coordinates rather than contributes, that formats information for upward consumption and runs recurring status meetings, is under pressure in nearly every large organization deploying AI in earnest. Chesky simply said it would disappear.
The question for Canadian HR leaders is not whether they agree. It is what they are going to do about it.
What Chesky actually said - and what he did not
The full context matters because the headline version - "people managers are finished" - is both accurate and incomplete.
READ MORE: AI is an enabler - it can't replace the human in HR, says CPO
Chesky's preferred model is what he calls a "hybrid people manager" or "manager IC" - an individual contributor who also leads people, whose authority comes from expertise and proximity to the work. His example was Jony Ive, former Apple chief design officer: a leader whose influence derived from direct engagement with the product, not from a position in an org chart.
He pointed to Airbnb as a live example of the transition in action. Design and engineering leaders are "going back to coding or using Claude Code," Anthropic's AI coding assistant. AI now generates nearly 60 per cent of Airbnb's code. The managers thriving, he said, are those in the work - not those supervising it from a remove.
He was careful to distinguish the role becoming less valuable from the person becoming redundant. "The companies that are prepared to change and transform are the companies that are going to benefit from AI. And if you do not change, you are going to be disrupted."
A chorus, not a solo
Chesky would be easier to dismiss if he were alone. Jack Dorsey of Block has written that "there is no need for a permanent middle management layer." Coinbase, cutting 14 per cent of its workforce, is simultaneously flattening to five organizational layers below the CEO, with every leader required to be a "strong and active contributor." Gartner predicts that by 2026, 20 per cent of organizations will use AI to eliminate more than half of their middle management positions.
The Canadian data adds a specific dimension. Only 12 per cent of Canadian businesses reported using AI in their operations as of mid-2025 - one of the lowest adoption rates among comparable economies. Yet 68 per cent of workers say they are ready to embrace AI, and 37 per cent say they would consider leaving their job if AI training is not provided. That is a workforce that is ahead of its organizations on intent and being left behind by its managers on execution.
Research from Randstad Canada found that 83 per cent of employers are moving toward skills-based models, reflecting a recognition that adaptability and human judgment are the durable skills as automation expands. That shift is precisely what Chesky is describing at the management level: the future manager is defined by domain skills and judgment, not by the act of supervising others.
The research says Chesky is partly right - and the "partly" matters
The evidence on middle management and AI is more nuanced than the tech CEO chorus suggests, and Canadian HR leaders should hold both truths simultaneously.
In Korn Ferry's Workforce 2025 survey of 15,000 professionals, 41 per cent of employees said their organisation had cut management layers - with middle managers "often the first in line" when companies cut labour costs. That confirms Chesky's diagnosis at the trend level.
But the mechanism matters. HRD Canada has reported that AI tools in IBM's HR system have actually reduced managerial errors and freed up HR professionals to allocate more time to strategic functions like talent and leadership development. The finding is not that managers become unnecessary when AI handles their administrative load. It is that they become more capable - if the organization invests in making that transition deliberate rather than accidental.
READ MORE: Can AI help put the human back in HR?
HRD Canada's CPO interview at TD Synnex captured the necessary counter-argument: "Judgment, empathy, and understanding context are areas where AI falls short. For instance, while AI can analyse data to identify high-potential candidates, it can't assess cultural fit or leadership potential during nuanced conversations." Strip management of its human judgment function without replacing it, and the result is not a leaner, more agile team. It is a workforce that has lost the connective tissue between strategy and execution.
Meeting technologies, told HRD Canada, are giving managers vital clues about engagement and attention in virtual settings - helping them be "more effective coaches committed to bringing the best out in their people." That is the player-coach model Chesky advocates: AI augmenting the human manager's capacity for connection, not replacing the need for connection altogether.
What HR must own
Chesky is describing a destination. Getting there belongs to HR. Several practical imperatives follow.
The leadership development pipeline needs to be rebuilt around the player-coach model. Succession frameworks that identify future leaders based on their people management skills alone are calibrated for a world that is ending. The capability frameworks Canadian organizations are investing in need to weight domain expertise, AI fluency, and direct contribution alongside - not beneath - people leadership.
The 37 per cent of workers who would leave without AI training are telling HR something important: the manager who cannot model AI use, answer questions about it, and create space for experimentation is no longer adequate. That is not a technology department problem. It is a management development problem.
And the ambiguity that Chesky is exploiting - "pure people managers have no value" - is not a reason for HR to stay silent. HRD Canada has reported that organizations pulling ahead are those that "blend technology with a human touch, founded on trust". That blend requires a human in the middle of the technology. The manager is not the problem. An underdeveloped manager is the problem. And that is fixable.
The meeting manager is not over. The meeting-only manager is. That distinction is the work.
